Cornell University Law School Social Science and Law

Jury Size: Less in not More

Compiled by Evan Moore and Tali Panken

INTRODUCTION: George Zimmerman and the Panel of Six

Last year, national news organizations picked up the sensational and racially charged story of the shooting of Trayvon Martin by George Zimmerman, a neighborhood watch volunteer who claimed immunity under Florida’s Stand Your Ground law. Coverage of jury selection and deliberation during Zimmerman’s murder trial frequently mentioned that the jury was comprised of just six individuals, all women (five white, one black). Many commentators and observers were surprised that Florida criminal trial juries were composed of just six members, instead of the traditional twelve.ⁱ Florida is an outlier since most states do not use juries of six for criminal trials; six is now the norm in civil trials, however, in the majority of states.

Six-person juries are the product of a 1970 Supreme Court decision, Williams v. Florida (399 U.S. 78), that is now severely criticized both for the empirical studies on which the Court relied and the Court’s analysis of those studies. In fact, modern empirical research in sampling theory and group dynamics suggests that six-person juries are less representative of their respective communities than twelve-person juries and less deliberative and thoughtful than their larger cousins.

CASE LAW

DEPARTURE FROM 700 YEARS OF PRECEDENT: Williams v. Floridaⁱⁱ and the Jury of Six

During the 1960s, court reform movements pressed to both increase the efficiency and decrease the cost of court proceedings. One primary target was the traditional jury of twelve peers, and by the late 1960s Florida passed a law that provided for juries of six in civil and criminal trials.

Williams was convicted of armed robbery in Florida by a jury of six and sentenced to life in prison. On appeal, he argued that the Sixth Amendment provided for jury trial according to its characteristics under English and American common law, consisting of twelve peers.

The Supreme Court affirmed Williams’ conviction. It first determined that modern juries were not bound to their common law form, and then adopted a functional equivalence test for any reduction in jury size from the traditional twelve. Relying on little more than intuition, the Court found that a six-person jury was unlikely to be any less representative of the community than a twelve-person one by a sample size reduction of 50% close to the lower bound. Furthermore, it conflated empirical and non-empirical studies to find that the jury’s deliberation and fact-finding abilities would not be significantly affected in six-person form.

The non-empirical research the Court cited in support of six-person juries included:

• Six-Member Juries Tried in Massachusetts District Court, 42 J.Am.Jud.Soc. 136 (1958);
• Note, On Instructing Deadlocked Juries, 78 Yale L.J. 100, 108 and n. 30 (and authorities cited), 110-111 (1968); and
• C. Joiner, Civil Justice and the Jury 31, 83 (1962) (concluding that the deliberative process should be the same in either the six- or 12-man jury).
  
  

FIVE IS NOT ENOUGH: Ballew

A. Summary

In 1978, the Supreme Court further reconsidered the constitutional requirements of jury sizein Ballew v. Georgia. Ballew was charged for two counts of misdemeanors for distributing obscene materials in violation of Georgia law. He went to trial in the Criminal Court of Fulton County, Georgia, where the court tried misdemeanor cases before five-person juries. Ballew’s request for a twelve-person jury was denied, and he was found guilty by the trial court jury. On appeal, Ballew raised a Sixth Amendment challenge against the five-person jury, but he was unsuccessful. The Supreme Court granted certiorari and considered whether a five-person jury in a state criminal trial was valid under the Sixth and Fourteenth Amendments.

While the Court upheld the constitutionality of a six-person jury in Williams, it purposefully did not address whether a jury with fewer than six individuals was constitutional.In Ballew, the Supreme Court was determined to answer whether a jury with fewer than six individuals "inhibite[ed] the functioning of the jury as an institution to a significant degree, and, if so, whether any state interest counterbalances and justifies this disruption." After reviewing contemporary empirical research, the Court in Ballew held that a jury with fewer than six members seriously impaired the “purpose and functioning” of a jury in criminal trials.

B. Empirical Data presented by Ballew

First, the Court looked at contemporary research demonstrating that smaller juries were less likely to facilitate effective group deliberation.

• A study by Thomas and Fink showed that group size was an important variable in the qualities of group deliberation. After reviewing 31 studies of small groups where group size was the independent variable, they found that there were no conditions under which smaller groups showed superior skills in group performance and group productivity. (Thomas & Fink, Effects of Group Size, 60 Psych.Bull.371, 373 (1963) available at http://psycnet.apa.org/journals/bul/60/4/371/).
• In smaller groups, members are less likely to make critical contributions to solve a given problem. (Faust, Group versus Individual Problem-Solving , 59  J.Ab. & Soc.Psych 68, 71 (1959) at http://psycnet.apa.org/journals/abn/59/1/68/)
• In smaller groups, members are less likely to overcome their biases and obtain an accurate result. (Barlund, A Comparative Study of Individual, Majority, and Group Judgment, 58 J.Ab & Soc.Psych. 55, 59 (1959).
  
  

Second, the Court found that smaller juries were less likely to reach accurate results.

• Nagel & Neef- Statistical studies show that the risk of Type 1 errors (convicting an innocent person) increase as the size of the jury decreases. Conversely, the risk of Type 2 errors (the error of not convicting a guilty person) increases as the size of the jury increases. Nagel and Neef determined that the optimal size for reducing both errors should be a function of the two types of errors, weighing the Type 1 error as 10 times more significant that a Type 2 error. They concluded that the optimal jury size was 6-8 members. (Nagel & Neef, Deductive Modeling to Determine an Optimum Jury Size and Fraction Require to Convict, 1975 Wash.U.L.Q. 933)
  
  

Third, the likelihood of a hung jury decreases as the juries reduce in size; this disadvantages the defense. Hung juries generally only occur where one or two jurors are unconvinced of guilt.

• Zeisel- Zeisel estimated that the number of hung juries in 6-person juries (2.4%) is half the amount of hung juries in 12-person juries (5%).(Zeisel, . . . And Then There Were None: The Diminution of the Federal Jury, 38 U.Chi.L.Rev. 710 (1972)).

Fourth, the Court was troubled by the fact that jury members from minority groups are less likely to be represented in smaller groups:

"Thus, if a minority group constitutes 10% of the community, 53.1% of randomly selected six-member juries could be expected to have no minority representative among their members." (citing Michael Saks, Ignorance of Science is No Excuse (1974))

C. Commentary

Justice Powell, in his concurrence, questioned whether Blackman’s use of statistical research was necessary. He also wrote, "Neither the validity nor the methodology employed by the studies cited was subjected to the traditional testing mechanisms of the adversary process."

In Gonzalez v. Florida, the Calebresi Amici Brief raises several issues with the Ballew decision. Ballew abandoned the functional equivalence test used in Williams in favor of a bright-line rule that six-person juries were constitutional, but five-person juries were unconstitutional. However, this bright-line rule seems arbitrary given that the Court relied on studies showing that the six-person juries perform worse than twelve-personjuries. Here are some of the studies comparing six-and twelve-person juries which were cited by the Court in Ballew:

• Davis, Kerr, Atkins, Holt, & Mech, The Decision Processes of 6-and 12-Person Mock Juries Assigned Unanimous and Two-Thirds Majority Rules, 32 J. of Personality & Soc. Psych. 1 (1975)
• Friedman, Trial by Jury: Criteria for Convictions, Jury Size and Type I and Type II Errors, 26-2 Am.Stat. 21 (Apr.1972)
• Pabst, Statistical Studies of the Costs of Six-Man versus Twelve-Man Juries, 14 Wm. & Mary L.Rev 326 (1972)
• Note, Six-Member and Twelve-Member Juries: An Empirical Study of Trial Results, 6 U.Mich.J.L. Reform 671 (1973)
• Note, An Empirical Study of Six-and Twelve-Member Jury Decision-Making Processes, 6 U.Mich.J.L. Reform 712 (1973)
  
  

Critics have wondered why Blackmun did not overturn the holding in Williams given the social science evidence that six-person juries do not deliberate as well as twelve-person juries. Perhaps the social scientists who conducted these studies with six-person juries were naïve to believe that the Supreme Court would overturn a constitutional holding based on empirical evidence.

FURTHER RESEARCH

REPUBLIC OF THE COURTROOM: Jury Size and Population Representativeness

One of the persistent concerns with juries, traceable back to the Magna Carta and the Star Chamber, is that they be composed of one’s peers, or in the language of the Sixth Amendment, that they be impartial. As Shari Diamond puts it, the jury is supposed to be a representative sample of the community.ⁱⁱⁱ Sampling theory suggests, and Lempert confirms in his "Uncovering ‘Nondiscernible Differences: Empirical Research and the Jury Size Cases"^iv, that larger juries are more likely to include individuals of diverse backgrounds, beliefs, and experiences than smaller ones. The question becomes one of degree; specifically, how much less representative of a community population is a jury of six likely to be than a jury of twelve?

Many observers expressed surprise at the homogeneity of the Zimmerman jury.^v While there was some consternation that the jury only included one minority member, Seminole County, FL, is just 11% African American. The chance of at least one African American being on the jury of six was 51%, or roughly 1 in 2.^vi  The jury’s greater statistical oddity was that all six members were female, of which there was less than a 2% chance. Modern empirical sampling theory demonstrates that juries of six members are much more susceptible to outlier effects, such as an all-female jury. The chart below shows a normally distributed, randomly drawn 1000-jury sample and the deviation from the mean of those juries in terms of the number of women. The less peaked distribution for six-member juries is a visual depiction of samples with a greater chance of outliers than those for twelve-member juries, which cluster around the female population mean.

Six-member juries can produce clearly visible effects that would be remotely unlikely in a twelve-member jury. In the Zimmerman case, recall that there was less than a 2% chance that all jurors would be female. In a twelve-member jury, however, that chance drops to less than .04%, from 1-in-51 odds to 1-in-2,558. Furthermore, the likelihood of an African American juror climbs from 51% to 76%. Other minority groups present in the community also demonstrate much greater chance of representation in the jury as its size increases, and it may be presumed, while it is much more difficult to categorize than race or ethnicity, that a broad array of experiences and backgrounds are also better represented.^vii  

Finally, representativeness of a community may include its composition, but it may also include reflection of the community’s views. Michael Saks has studied view representativeness at length, and he finds that as jury size decreases, verdict and award unpredictability increase. For every halving of the sample size (jury), variability in outcome will increase by 41%. Verdict and award predictability is significant if we assume that the overall population has a distinct view of the correct verdict and award amount. Since twelve-member juries tend to reduce outlier awards and conform more strictly to the mean population verdict decision and award than six-member juries, twelve-member juries are more representative of their community in the trial outcome reached. For reference, the visual effect of jury size on a normal distribution of awards is shown below, although the actual curve varies somewhat from the standard normal distribution.^viii

JURY DELIBERATION: A Function of Number
           
The empirical studies used by the Court in Ballew represents foundational research on the topic of jury size and jury deliberation. In 1997, Michael Saks (whose empirical work influenced the Ballew opinion) and Mollie Marti conducted a meta-analysis of 17 studies that examined the differences between six- and twelve-person juries. For the most part, this study confirmed the empirical work presented in Ballew. Saks and Marti found that larger juries were more likely to contain members of minority groups, have a hung jury, and accurately recall trial testimony (an indicator of juror accuracy). 

Modern research on this topic has continued to examine how having juries with diverse racial compositions affects jury deliberation. Some researchers hypothesized that minority jurors would be less likely to participate in less-diverse groups, but at least one study showed that diversity did not affect the participation levels of minority jurors.  Cornwell, York, and Hans analyzed data from 2,189 jurors on criminal cases in four jurisdictions to consider what conditions influence participation in jury deliberations. They suggest that "full participation by jurors from diverse backgrounds" is beneficial for jury fact-finding. To gather statistics about juror participation, the researchers relied on jurors’ self reports about their participation.  Results showed that Black jurors participated at high levels regardless of the group’s diversity.

Sommers found that the racial composition of juries affected the deliberative content. Coders evaluated the deliberative content of the mock juries and found that White jury members were more likely to raise novel facts in diverse jury deliberations than all-White jury deliberations. Also, uncorrected factual errors were less frequent in diverse groups than in all-White groups. Black participants were more likely to raise novel, race-related issues than White participants in diverse groups. Sommers concluded that diversity affected White participants’ informational processes and how they interpreted and weighted the evidence. (Sommers, On Racial Diversity and Group Decision Making: Identifying Multiple Effects of Racial Composition on Jury Deliberation (2006)).

Measure	Diverse Group	All-White group
Deliberation length (in min)	50.67	38.49
No. of case facts discussed	30.48	25.93
No. of uncorrected inaccurate statements	1.36	2.49
No. of race-related issues raised	3.79	2.07

Larger group sizes mean that there is a larger representative sample. In addition to increasing the likelihood of diversity, a larger jury is more likely to have a wider range of talents, skills for performing specialized tasks, and knowledge. As group sizes increase, there is an increasing organization and division of labor amongst the members. (Thomas & Fink, Effects of Group Size (1963)). Also, Hans conducted an experiment where she asked judges and juries to watch a mock trial in which mtDNA was at issue and then take a quiz testing how well they understood the scientific evidence. She found that college-educated jurors were more competent at fact-finding than jurors who were not college educated. Jurors outperformed judges on a question about the maternal heritage of mtDNA, which highlights the value of juror deliberation. (Hans, 44,Judges, Juries, and Scientific Evidence (2007)). If larger juries make it more likely that there will be a juror with a background in science, one can infer that larger juries are more likely to be capable of handling scientific evidence presented at trial. This is supported by an experiment with mock jurors which found that twelve-person juries recalled more probative evidence and relied less on non-probative evidence than six-person juries. (Horowitz and Bordens, The Effects of Jury Size, Evidence Complexity, and Note Taking on Jury Process and Performance in a Civil Trial (2002)).
           
Additionally, it should be noted that efficiency is one justification for using smaller juries. Most research indicates that, on average, larger juries spend a longer time in deliberation than smaller juries. Saks and Marti report that the unweighted mean length of deliberation time was 53 minutes for small juries and 70 minutes for large juries. While this finding supports the efficiency justification for smaller juries, it is possible that the greater deliberation time indicates a more thoughtful and collaborative review of the trial testimonies and evidence. Smaller juries thus might save a marginal amount of time but result in reduced deliberation. Below is a chart of the mean length of deliberation time for small and large juries in different studies:

Endnote Resources:

ⁱ Florida death penalty cases use a full 12-member jury.

ⁱⁱ 399 U.S. 78, 1970.

ⁱⁱⁱ Diamond, Shari. "Zimmerman trial: Time to reconsider six-member jury", Miami Herald. http://www.miamiherald.com/2013/07/14/3497719/zimmerman-trial-time-to-reconsider.html

^iv 73 Mich. L. Rev. 643.

^v See, e.g., http://www.theatlantic.com/national/archive/2013/06/why-the-george-zimmerman-trials-all-female-jury-is-news/277103/

^vi Probabilities exclude, for simplicity’s sake, jury selection and peremptory challenge effects. Probabilities and distributions were calculated for this project.

^vii Lempert. "Nondiscernable Differences".

^viii Saks, Michael. "The smaller the jury, the greater the unpredictability." 79 Judicature 263.